Case report

Pericardial graft vs. host disease in a patientwith myelodysplastic syndrome followingperipheral blood stem cell transplantation

The clinical features of chronic graft vs. host disease(GVHD) resemble those of several autoimmunediseases (1). Isolated pericardial effusion is occa-sionally observed in patients following allogenicstem cell transplantation, and may be due to adverseeffects of intensive chemotherapy or total bodyirradiation (TBI) during the conditioning phase (2),or may be caused by acute GVHD within 100 dpost-transplantation (3). However, pericardial effu-sion has rarely been reported as a manifestation ofchronic GVHD developing later than 100 d post-transplantation.We report a patient with pericardialeffusion caused by chronic GVHD, and describecytokine profiles, which were closely associated withthe clinical course of pericardial effusion.

Case report

A 42-year-old woman with myelodysplastic syn-drome (refractory anaemia with excess of blasts;RAEB) was referred to our hospital in October2000. Cytogenetic analysis revealed a karyotype of46, XX, i(17)(q10). In July 2001, the patientreceived allogenic peripheral blood stem cell trans-plantation from her HLA-identical sister followinga preparative regimen consisting of cyclophospha-mide (total dose, 120 mg/kg) and fractionated totalbody irradiation (total, 12 Gy). GVHD prophy-laxis consisted of continuous infusion of cyclo-sporin A (CsA) and short-term methotrexate.Successful donor cell engraftment, defined as a

Saito Y, Matsushima T, Doki N, Tsumita Y, Takizawa M, YokohamaA, Handa H, Tsukamoto N, Karasawa M, Murakami H, Nojima Y.Pericardial graft vs. host disease in a patient with myelodysplasticsyndrome following peripheral blood stem cell transplantation.Eur J Haematol 2005: 75: 65–67. � Blackwell Munksgaard 2005.

Abstract: A patient with myelodysplastic syndrome developed peri-cardial effusion 20 month after allogenic peripheral blood stem celltransplantation. Sclerotic and erythematous skin lesions were observedover the face and extremities, and a diagnosis of chronic graft vs. hostdisease (GVHD) was made based on skin biopsy findings. Pericardialfluid contained numerous CD8+/HLA-DR+ lymphocytes, but noleukaemic cells. Tumour necrosis factor alpha (TNFa) and soluble Fas(sFas) levels were highly elevated in both the effusion and serum. Thepatient was treated with methylprednisolone and tacrolimus. SkinGVHD improved rapidly associated with resolution of pericardialeffusion and reductions in cytokine levels. We concluded that pericardialeffusion was due to pericarditis and was a manifestation of chronicGVHD in this patient, and that cytotoxic lymphocytes and specificcytokines played significant roles.

Yasuyuki Saito1, TakafumiMatsushima1, Noriko Doki1, YukiTsumita1, Makiko Takizawa1,Akihiko Yokohama1, HiroshiHanda2, Norifumi Tsukamoto1,Masamitsu Karasawa3, HirokazuMurakami2, Yoshihisa Nojima11Department of Medicine and Clinical Science, GunmaUniversity Graduate School of Medicine, Gunma, Japan;2School of Health Sciences Faculty of Medicine, GunmaUniversity, Gunma, Japan; 3Division of BloodTransfusion Service, Gunma University Hospital, Gunma,Japan

Key words: cytokines; graft vs. host disease;pericarditis; stem cell transplantation

Correspondence: Yasuyuki Saito MD, Depertmentof Medicine and Clinical Science, Gunma UniversityGraduate School of Medicine,3-39-22 Showa-machi,Maebashi, Gunma 371-8511, JapanTel: +81-27-220-8166Fax: +81-27-220-8173e-mail: yass-gnm@umin.ac.jp

Accepted for publication 18 January 2005

Eur J Haematol 2005: 75: 65–67All rights reserved

Copyright � Blackwell Munksgaard 2005

EUROPEANJOURNAL OF HAEMATOLOGY

65

peripheral neutrophil count >0.5 · 109/L, wasobserved on day +14. Acute GVHD (grade II)developed on day +20. Tacrolimus and predniso-lone (PSL) were substituted for CsA to treat acuteGVHD. Although, GVHD was well controlled, shedeveloped pure red cell aplasia (PRCA) witherythroblastopenia in the bone marrow, requiringrepeated blood transfusions. She underwent plasmaexchange four times to treat PRCA in May 2002,and had been free of transfusion thereafter. Nine-teen months after transplantation, tacrolimus wasdiscontinued because she no longer showed obviousactive GVHD and there were no abnormal findingson chest X-ray or echocardiography.

One month after discontinuation of tacrolimus,the patient gained 4 kg in body weight. Facialoedema developed, which was resistant to diuretics.She was readmitted to our hospital in March 2003,and chest X-ray revealed cardiomegaly (cardio-thoracic ratio, 51%). Echocardiography showedpericardial effusion. The pericardial fluid wasdrained; a haemorrhagic exudative fluid wasobtained, which contained numerous infiltratingmononuclear cells without malignant cells. Cellcount in the effusion was 0.8 · 109/L with 30%lymphocytes and 70% macrophages. Most infiltra-ting lymphocytes were positive for CD8 and HLA-DR antigens. There were no laboratory findingssuggestive of viral or bacterial or fungal infection inthe blood or pericardial effusion. There was noevidence of infection by Epstein-Barr virus orcytomegalovirus based on the negative findings onpolymerase chain reaction or antigenemia assay,respectively. Thyroid hormone level was within thenormal range. Anti-nuclear antibody was positiveat 320· with a nucleolar pattern. Bone marrowexamination revealed normal karyotype and exclu-ded relapse. On day 10 after admission, her weighthad increased by a further 12 kg. Erythematousand sclerotic skin changes developed on her face,upper arms and lower legs. Histological findings onskin biopsy were compatible with chromic GVHD.She was treated with tacrolimus (2 mg/d) andmethylprednisolone pulse therapy (1 g/d for threeconsecutive days). Her skin involvement improvedrapidly and cardio-thoracic ratio decreased to 43%1 month after treatment. Echocardiography con-firmed resolution of pericardial effusion, althoughsmall amounts of fluid were still detected.

We investigated cytokine profiles in serum andpericardial fluid by ELISAs for IL-2 (Otsuka AssayLaboratory, Tokushima, Japan), IL-4 (R & DSystems, Minneapolis, MN, USA), IL-10, TNFa(BioSource Europe, Nivelles, Belgium), IFNc (Im-munotech, Marseilles, France) and sFas (Medical& Biological Laboratories, Nagoya, Japan)(Table 1). While IL-2, IL-4, IL-10 and IFNc were

undetectable, sFas and TNFa levels in serum andpericardial fluid were elevated significantly. Thelevels of sFas and TNFa decreased concomit-ant with clinical improvement, although theyremained above the reference ranges 3 month afteradmission.

Discussion

We described isolated pericardial effusion in apatient post-transplantation. The concomitantdevelopment of histologically confirmed chronicGVHD and the good response to immunosuppres-sive therapy suggested that pericardial effusion wasdue to pericarditis and was a manifestation ofchronic GVHD in this patient.

Development of pericardial effusion during thelate post-transplantation period has been reportedin six patients: four had pericardial effusion withdevelopment of extensive chronic GVHD andsimultaneous pleural and/or peritoneal effusion(4–7), while in the other two patients, the causesof pericardial effusion were recurrence of leukaemia(8) and Streptococcus pneumonia infection (9). Theclinical features of chronic GVHD resemble thoseof autoimmune disorders (1). Most patients withchronic GVHD show B-cell dysregulation withautoantibodies to cell-surface and intracellularantigens (6). All four cases with pericardial chronicGVHD showed high ANA titres. Retrospectively,we found that ANA had preceded pericarditis by3 month in our patient.

In addition to humeral abnormalities, donor Tcells reacting with self-components of recipientshave been suggested to play a pivotal role in thepathogenesis of chronic GVHD. Infiltratinglymphocytes expressed both CD8 and HLA-DRantigens in our patient. Likewise, Ueda et al.reported that lymphocytes in pericardial effusionfrom patients with chronic GVHD were predom-inantly CD8- and HLA-DR-positive (5). These

Table 1. Cytokine levels in pericardial fluid and serum

Cytokines Unit

Pericardial fluid Serum

OA OA 1M 3M

IL-2 pg/mL <50 <50 – –IL-4 pg/mL <15 <15 – –IL-10 pg/mL <8 <8 – –IFNc IU/mL <0.4 <0.4 – –TNFa pg/mL 62 59 10 19sFas ng/mL 4.05 5.31 3.59 3.93

OA: on admission; IL-2: interleukin 2 (reference range in serum, <50 pg/mL); IL-4:interleukin 4 (reference range in serum, <15 pg/mL); IL-10: interleukin 10 (referencerange in serum, <8 pg/mL); IFNc, interferon gamma (reference range in serum,<0.4 IU/mL); TNFa, tumour necrosis factor alpha (reference range in serum,<6 pg/mL); sFas: soluble Fas (reference range in serum, 0.33–1.53 ng/mL).

Saito et al.

66

findings suggest that cytotoxic T cells play a role ininducing pericarditis in chronic GVHD.TNFa and sFas were significantly elevated in our

patient’s serum and pericardial fluid, while IFNc,IL-2, IL-4 and IL-10 were undetectable. Thiscytokine profile differs from those in pericardialeffusion caused by bacterial infections, where levelsof IFNc, IL-2, IL-10 and TNFa are high (10).Chronic GVHD remains the most significant

complication after allogenic haematopoietic stemcell transplantation although its pathophysiology ispoorly understood. Pericardial chronic GVHD is arare but fatal complication, and is often difficult todistinguish from infection. Examination of peri-cardial fluids, especially lymphocyte and cytokineprofiles, may provide useful information for diag-nosis as well as for monitoring disease activity.

References

1. Ratanatharathorn V, Ayash L, Lazarus HM, Fu J,Uberti JP. Chronic graft-versus-host disease: clinicalmanifestation and therapy. Bone Marrow Transplant2001;28:121–129.

2. Veys PA, McAvinchey R, Rothman MT, Mair GH,Newland AC. Pericardial effusion following conditioningfor bone marrow transplantation in acute leukaemia. BoneMarrow Transplant 1987;2:213–216.

3. Seber A, Khan SP, Kersey JH. Unexplained effusions:association with allogeneic bone marrow transplantationand acute or chronic graft-versus-host disease. BoneMarrow Transplant 1996;17:207–211.

4. Toren A, Nagler A. Massive pericardial effusion com-plicating the course of chronic graft-versus-host disease(cGVHD) in a child with acute lymphoblastic leukemiafollowing allogeneic bone marrow transplantation. BoneMarrow Transplant 1997;20:805–807.

5. Ueda T, Manabe A, Kikuchi A et al. Massive pericardialand pleural effusion with anasarca following allogeneicbone marrow transplantation. Int J Hematol 2000;71:394–397.

6. Silberstein L, Davies A, Kelsey S et al. Myositis,polyserositis with a large pericardial effusion andconstrictive pericarditis as manifestations of chronicgraft-versus-host disease after non-myeloablative periph-eral stem cell transplantation and subsequent donorlymphocyte infusion. Bone Marrow Transplant2001;27:231–233.

7. Law L, Tuscano J, Wun T, Ahlberg K, Richman C.

Filgrastim treatment of acute myelogenous leukemia (M7)relapse after allogeneic peripheral stem cell transplantationresulting in both graft-versus-leukemia effect with cyto-genetic remission and chronic graft-versus-host diseasemanifesting as polyserositis and subsequent bronchiolitisobliterans with organizing pneumonia. Int J Hematol2002;76:360–364.

8. Einsele H, Ehninger G, Vallbracht A et al. Isolatedpericardial relapse following allogeneic bone marrowtransplantation for acute myelogenous leukemia. BoneMarrow Transplant 1989;4:323–325.

9. Perez Retortillo JA, Marco F, Richard C, et al.Pneumococcal pericarditis with cardiac tamponade in apatient with chronic graft-versus-host disease. BoneMarrow Transplant 1998;21:299–300.

10. Burgess LJ, Reuter H, Carstens ME, Taljaard JJ,Doubell AF. Cytokine production in patients withtuberculous pericarditis. Int J Tuberc Lung Dis2002;6:439–446.

Pericardial GVHD after PBSCT

67